Runway, the AI video company, has formally denied that it got around YouTube’s technical protections to collect creators’ videos for training. In an answer filed on 13 November 2026 in federal court in Manhattan, it denies the core allegations and lists 15 defenses, from authorization to free speech.

The filing comes two weeks after Judge Lewis Kaplan denied Runway’s motion to dismiss on 30 October. The case now moves toward evidence. Nothing has been decided about what Runway actually did.

What the creators claim

The plaintiffs are Ace Cam, which trades as Random Golf Club, and Businessing LLC, suing for a proposed class of YouTube creators. According to the opinion, Businessing alleges at least 1,200 of its videos were scraped and Random Golf Club at least 390.

The suit makes one claim, under the anti-circumvention rule of the Digital Millennium Copyright Act (DMCA), not a copyright infringement claim. Section 1201(a)(1)(A) says: “No person shall circumvent a technological measure that effectively controls access to a work protected under this title.” Section 1201(a)(3)(A) defines circumvent to include “otherwise to avoid, bypass, remove, deactivate, or impair a technological measure.”

The creators point to five YouTube technological protection measures (TPMs), including a rolling cipher, rate limits, CAPTCHA tests and session-bound links. They allege Runway used descrambling tools such as yt-dlp and rotated IP addresses to avoid being blocked.

What the judge said

Runway argued that anyone can watch YouTube for free, so nothing was locked. Judge Kaplan disagreed, at a stage where the court assumes the complaint’s facts are true. “The public nature of content that has been uploaded to YouTube does not render it unprotected because the public’s viewership of the content necessarily is subject to YouTube’s TPMs.”

He also held that creators need not install the protections themselves. He cited an earlier 2026 ruling from the same court in Reddit’s case against SerpApi, a link the SDNY Blog summary also draws.

What Runway’s answer says

Runway admits it “uses data to train the Models,” a list that runs from Gen-2 to Gen-4.5. It admits it “has not publicly disclosed the full composition of the data corpus,” which it calls confidential. It denies that YouTube’s measures are access controls and that IP rotation is circumvention.

Three defenses stand out. One invokes Section 1203(c)(5), which lets a court cut damages if the violator “was not aware and had no reason to believe” it broke the law. Another calls any per-act statutory award unconstitutional in this case. A third says the creators’ reading would “restrict virtually all speech on the internet.”

Follow the money and the damages defense makes sense. Section 1203(c)(3)(A) sets statutory damages at “not less than $200 or more than $2,500 per act of circumvention.” Across a class, that adds up fast. An AI litigation blog post, republished by Web Pulse, noted that listing a defense does not mean Runway will press it. We found no trial date in either filing.

The case sits apart from fights over whether training is fair use, such as the ROSS rehearing request. It asks how the videos were collected. Scraping also turns up in platform contracts, as when Meta paused business terms that would have let it scrape linked sites.

What to do this week

If your team or agency pulls YouTube video with download scripts, proxy pools or a scraping vendor, list those jobs now. A US judge has found that claims of dodging rate limits or defeating a cipher plausibly describe circumvention, even for free public videos.

Ask any vendor that supplies video data how it was collected, and get the answer in writing. Use footage you own, license or have written permission for. This is one US case at an early stage, so check with a lawyer before changing anything you rely on.

Sources

Martin Scott

Martin Scott is the editor for business and policy at AI Observer, covering startups and deals, regulation, health and science, and AI agency profiles. He follows the money and the legal text, and separates what a rule says from what people claim it says. His style is analytical and plain-spoken.

Spotted a mistake? Tell us on the corrections page.

Conversation

Every comment is reviewed before it appears, so there can be a short delay.